Developing evidence strongly suggests that bipolar disorder involves localized abnormalities in specific brain structures that participate in fronto-limbic-subcortical circuits regulating mood. The specific structures include the prefrontal cortex, anterior cingulate cortex, thalamus, basal ganglia, amygdala hippocampus, and cerebellum. There is clearly a need for systematic study of these structures in bipolar disorder to determine the magnitude and extent of regional neuropathology, which may be too subtle to reliably detect by present quantitative MRI measurements of tissue volume alterations. Subtle neuronal loss or damage has been shown to be readily detectable by proton magnetic resonance spectroscopy (1 H MRS I) measures of N-acetylaspartate (NAA), a neuronal and axonal marker and a reliable indicator of neuronal integrity. Therefore, this proposal will utilize high resolution MRI together with 'H MRSI to determine if there is reduced NAA, with or without tissue volume loss, in the dorsolateral prefrontal cortex, orbitofrontal cortex, anterior cingulate cortex, caudate, putamen, hippocampus, and cerebellar vermis in bipolar disorder. MRI derived tissue volumes will be utilized together with NAA and choline measures to make preliminary inferences about regional neuropathologic changes such as neuronal dysfunction, neuronal loss (with or without gliosis), developmental hypoplasia, and synaptic/dendritic pruning failure. Second, this proposal will examine NAA measures in normal appearing white matter (NAWM) and white matter signal hyperintensities (WMSH) of the prefrontal white, subcortical, and periventricular white matter regions. White matter NAA reductions will be interpreted as evidence for axonal loss or dysfunction and evidence for compromised integrity of white matter pathways. Third, this proposal will determine if there is increased NAA. with or without increased tissue volumes, in the thalamus and amygdala in bipolar disorder. Increased NAA and/or increased tissue volumes will be interpreted as evidence for possible increased neuronal number or interneuronal neuropil. Fourth, this proposal will determine if the magnitude and/or extent of regionally specific NAA alterations is different in bipolar I compared to bipolar II disorder. Finally, this proposal will determine if longer illness duration and greater illness severity (defined as a greater number of lifetime hospitalizations for mania or depression) result in more severe brain pathology as measured by the magnitude of regionally specific NAA alterations. Determining whether the neuropathology in different brain regions is a result of progressive damage to the brain over time will have important implications for treatment interventions in binpolar disorder.